Figure 111 Spatial variability of landfill CH4 emissions from an intermediate cover area at a southern California landfill using static chambers

Note: Note the high positive CH4 fluxes in close proximity (<5m) to negative fluxes (uptake of atmospheric CH4) Source: Unpublished authors' data address the spatial variability (Webster and Oliver, 1992), which presents challenges for whole-site analyses. The use of a surrogate variable (one that is easier to measure than surface chamber emissions itself) could provide a mechanism through co-kriging to arrive at improved emission estimates if reliable and robust relationships can be drawn between the surrogate variable(s) and the corresponding emissions at a site (for example Ishigaki et al, 2005). Currently, co-kriging methods require further validation.

Explanations for the 'hot spots' (Figure 11.1) include localized areas of thinner cover materials, leakages at interfaces (edge of landfill footprint, localized emissions around gas piping systems requiring maintenance), macropore formation (for example animal burrows or desiccation cracking), and differential settlement from waste inhomogeneities, slope configurations, infiltration characteristics, and variable decomposition rates. Typically at US sites, for example, locations of such sporadic emissions are identified during routine surface monitoring for CH4 concentrations and then mitigated by follow-up maintenance activities. Imaging of depressions using remote sensing (thermal imagery) has met with mixed success (for example Zilioli et al, 1992; Lewis et al, 2003). Whole-landfill CH4 emissions measurements reported from Europe, the US and South Africa, also relying predominately on chamber methods, ranged within about one order of magnitude, from approximately 0.1 to 1.0t CH4 ha-1 d-1 (Nozhevnikova et al, 1993; Hovde et al, 1995; Czepiel et al, 1996a; Borjesson, 1997; Mosher et al, 1999; Trégoures et al, 1999; Galle et al, 2001; Morris, 2001).

During the last five years, more comprehensive campaigns using multiple techniques at the same site and comparing those techniques across several sites have recently increased our understanding of landfill CH4 emission rates. Moreover, as discussed below, the deployment of several 'above-ground' techniques in tandem in recent studies have also highlighted the limitations of techniques that are designed for uniform terrain, relatively constant meteorology, and emission sources with less extreme temporal and spatial variability than landfills.

Several years of studies at three French sites involved field-scale measurements of the complete CH4 mass balance (Equation 15) for nine different landfill cells with varying cover materials and management practices (Spokas et al, 2006, and references cited therein). Figure 11.2 compares modelled landfill gas generation from the nine sites to both measured gas recovery and measured emissions (using both a tracer plume method and dynamic chambers). Note that there is a significant linear relationship between generation and recovery (for the eight sites that had active gas extraction) but no relationship between generation and emissions, with the residual emissions varying over several orders of magnitude. Through a combination of intensive field measurements, supporting laboratory studies, and modelling, high rates of CH4 recovery were documented at the French sites. For example, only about 1-2 per cent of the CH4 production is being emitted and about 97 per cent is being recovered with an active gas extraction system at Montreuil-sur-Barse in eastern France (near Troyes). At Lapouyade (near Bordeaux, southwestern France), a minimum of 94 per cent of the CH4 production was being recovered at two cells with engineered gas recovery, but for a cell without recovery, 92 per cent of the CH4 production was being emitted (Spokas et al, 2006). Thus, through a combination of intensive field measurements, supporting laboratory studies, and modelling, high rates of CH4 recovery can be documented at field

Figure 11.2 Comparison of (A) measured CH4 emissions and modelled CH4 generation; and (B) measured CH4 recovery and modelled CH4 generation for nine whole-landfill cells in northeastern, western and southwestern France

Note: There is a lack of agreement between measured emissions and modelled generation as opposed to the good agreement between measured recovery and modelled generation. Source: Based on data from Spokas et al (2006)

scale. Additional mass balance studies are needed to expand our understanding of field-scale processes in various climatic regimes under different management strategies.

A direct field comparison of five above-ground techniques was also completed at the Lapouyade landfill in southwestern France in October 2007 (Babillotte et al, 2008). In addition, the ability of the techniques to quantify flux rates from a blind CH4 release test on a non-landfill area was also evaluated during the same field campaign. The techniques were (1) vertical radial plume mapping (VRPM) and horizontal radial plume mapping (HRPM) by ARCADIS (US); (2) differential absorption lidar (DIAL) by the National Physics Laboratory (NPL; UK); (3) mobile and static plume methods by ECN (The Netherlands); (4) inverse modelling from vehicle-based spectroscopic measurements by INERIS (France); and (5) airborne laser methane assessment (ALMA), a helicopter-based spectroscopy method by PERGAM (Switzerland). The ALMA/PERGAM method does not, however, yield quantitative measurements of fluxes. In general, the NPL and ARCADIS methods measured emissions from the four cells individually, yielding global site emissions of 12.8 ± 2.9, and 25.2 ± 2.4g sec-1, respectively. The ECN CH4 reported whole-site dynamic and static plume results of 41 ± 17 and 83 ± 36g sec-1 respectively, while the INERIS method reported a whole-site flux of 167.2g sec-1. In addition, a blind test was conducted off the landfill footprint with a CH4 release of 0.5g sec-1 over an area of 100m2. The comparison among the techniques (in terms of per cent of actual release rate detected by each method) is as follows: NPL (-54 per cent), Arcadis (+78 per cent), ECN (+240 per cent to +360 per cent) and INERIS (+200 per cent to +300 per cent). Both of the Gaussian dispersion methods (ECN and INERIS) significantly overestimated emissions. These results indicated, first, the complexities of landfill emissions measurements (adjacent cells with variable emissions) and some systematic differences among the four measurement methods. In general, there is, as yet, no single method that is appropriate for universal and unambiguous deployment for field measurement of landfill CH4 emissions.

In late September and early October 2008, Veolia Environmental Services (VES) and Waste Management, Inc. (WMX) collaborated on a comprehensive field comparison campaign at two adjacent landfill sites in southeastern, Wisconsin, US: the WMX Metro site and the VES Emerald Park site (Babillotte et al, 2009). Methane emissions from multiple areas of both sites were measured using multiple techniques, including (1) static chambers by WMX, Florida State University, and Landfills +, Inc.; (2) a micrometeorological method (eddy covariance) by the Finnish Meteorological Institute; (3) VRPM by WMX (in collaboration with ARCADIS, US); (4) DIAL by the National Physics Laboratory (UK); and (5) a mobile plume method using portable Fourier transform infrared (FTIR) spectroscopy by the Swedish company FLUXSENSE, which collaborates with Chalmers University. In addition, the comparative performance of (3)-(5) were evaluated using a blind test with a series of controlled CH4 release rates on a non-landfill area adjacent to the two sites. The release area encompassed 40 X 40m2 within an overall area of 300 X 500m. This area had favourable terrain (flat, uniform) and meteorology (steady wind speed with no contribution from landfill sources). As of the writing of this chapter, only the controlled release results have been published (Babilotte et al, 2009). Four separate trials, each with constant release rates from one to three locations on the test area, were conducted during midday hours (9.00-14.00). All groups were encouraged to optimize their own data collection through technique-specific experimental designs and instrument placement. The release rates (normalized on an area basis) and the results are summarized in Table 11.1. In general, for this trial, the FLUXSENSE mobile plume/portable FTIR method had the most unambiguous results (within 20 per cent of the actual release rate with low standard deviations) by measuring a well-mixed plume about 450m downwind from the release area. However, all three methodologies reported fluxes within about 20-30 per cent of the actual release rate. A major issue associated with the VRPM, which systematically underestimated fluxes, was that the magnitude of the underestimation was a function of the distance between the source and the vertical measurement plane, suggesting that existing models for the area contributing to flux require further refinement. The DIAL method enabled definition of plume and source area but had difficulties with quantification at larger distances, possibly due to measurement interferences. Thus, at the current time, there is no single recommended above-ground technique for field measurement of whole-landfill CH4 emissions. All of these techniques are being refined by the various research groups for their application to complex area sources such as landfills. In general, because static chambers can quantity the variability of emissions across a particular type of cover material (including locations where negative fluxes, or uptake of atmospheric CH4, is occurring), their use in parallel with above-ground methods such as those described above is highly recommended.

Methane oxidation for landfill settings has been studied at scales ranging from laboratory batch studies to field-scale measurements. In the field (Figure 11.3), oxidation tends to be optimized at specific depths that vary seasonally and can be observed in soil gas profiles. For small-scale laboratory batch studies, maximum CH4 oxidation rates have ranged from 0.01 to 117pg CH4 g-1 h-1 in landfill cover soils with organic carbon contents of 1.2 to 30 per cent (w/dry w). Q10 values ranged from 1.9 to 5.2 over temperature ranges of 2-30°C. Optimum moisture contents were generally below 25 per cent (w/w). For laboratory column experiments simulating landfill cover environments, steady-state CH4 oxidation rates (aerial basis) ranged from 22 to 210g CH4 m-2 d-1 with fractional CH4 oxidation ranging from 15 to 97 per cent for studies conducted over 30 to >300 days. The use of compost or other highly organic soils in similar column experiments over 35-369 days tended to increase the fractional CH4 oxidation to a higher range of approximately 70-100 per cent. (Scheutz et al, 2009, and references cited therein). In recent work elucidating the limits and dynamics of CH4 oxidation in landfill cover soils using batch studies, soils that were pre-incubated with CH4 for 60 days and with soil moisture potential adjusted to 33kPa (field-holding capacity) exhibited consistently high oxidation rates of 112.1 to 644pg CH4 gsoil-1 d1 across all soil types studied (Spokas and Bogner, 2010). These results contrasted with parallel studies of the same soils without pre-incubation and moisture adjustment which exhibited oxidation rates ranging over four orders of magnitude, from 0.9 to 277pg CH4 g-1 day1. In the same study, the minimum soil moisture threshold for oxidation activity was estimated at approximately 1500kPa. Furthermore, indicating the coupled interaction of soil moisture and temperature, the threshold soil moisture potential for CH4 oxidation activity shifted to lower soil moisture potentials (higher moisture contents) at extreme temperatures. At <5°C, the minimum soil moisture threshold is approximately 300kPa. At the upper temperature limits of CH4 oxidation activity (>40°C) the minimum soil moisture threshold is approximately 50kPa.

Figure 11.3 Soil gas concentration profile for CH4 and O2 through final cover materials at a southern California landfill

Note: Inset graph shows detail for the 0-100cm depth for CH4 (note decrease at 25cm). Source: Unpublished authors' data

At field scale, stable carbon isotopic methods, which rely on the difference between the 813C of emitted CH4 and the 813C of unoxidized CH4 in the anaerobic zone, provide the most robust approach to date for the quantification of fractional CH4 oxidation, that is, the percentage of CH4 that is oxidized during transport through the landfill cover materials. Isotopic methods have been developed over the last decade and rely on the preference of methanotrophs for the isotope of smaller mass, 12C rather than 13C, according to one or more fractionation factor(s) dependent on soil properties and gaseous transport considerations (Liptay et al, 1998; Chanton et al, 1999; Chanton and Liptay, 2000). Methanotrophic bacteria will oxidize 12CH4 at a slightly more rapid rate than 13CH4. In general, the isotopic methods can be applied as follows: (1) at ground level, using static chambers and comparing the 813C of CH4 in the refuse (by sampling at gas recovery wells, gas collection headers, or deep gas probes) to the emitted CH4 collected in the chamber; (2) in the lower atmosphere, relying on a comparison between the 813C of atmospheric CH4 in an upwind transect and a downwind transect; and (3) below ground level, relying on soil gas profiles for CH4 and 813C. In general, 813C values for unoxidized CH4 in the anaerobic zone range from about -57 to -60; with oxidation, these values can undergo a positive shift to -35 or more (Bogner et al, 1996, and references cited therein). An additional possibility is to use combined 813C and 8D methods since the hydrogen isotope has a larger relative fractionation factor than does the carbon. Recent studies have addressed these issues as well as expanded methods, models and uncertainties for isotopic approaches to CH4 oxidation in landfill cover soils (De Visscher et al, 2004; Mahieu et al, 2006; Chanton et al, 2008).

It is expected that future methodological and modelling improvements will permit more robust quantification of incremental CH4 oxidation through landfill cover soils. In situations where static chamber measurements yield 'negative' fluxes, indicating that methanotrophs are capable of oxidizing all of the CH4 transported from the landfill below and also oxidizing additional CH4 out of the atmosphere, the stable carbon isotopic methods are not applicable. In these cases, the static chamber is quantifying the rate of atmospheric CH4 oxidation/uptake, and data should be reported accordingly. As discussed above, where fluxes are positive, isotopic measurements comparing the 813C for emitted CH4 compared to CH4 in the anaerobic zone can be used to quantify the fractional CH4 oxidation - typically, fractional CH4 oxidation is significantly higher than the 10 per cent allowed by IPCC national inventory methodologies for annual reporting (IPCC, 2006; Chanton et al, 2008), based on Czepial et al (1996b), as discussed in more detail below. A recent review reported that the average percentage of CH4 oxidized in cover soil was 35 ± 6 per cent across a variety of landfill sites around the globe (Chanton et al, 2009).

Some researchers have relied on CH4:CO2 ratios as an indicator of CH4 oxidation in landfill soil gas profiles. This approach is not recommended because, in addition to the production of CO2 via methanogenesis and CH4 oxidation, CO2 is also produced and consumed by many other subsurface and

0 0

Post a comment